Copy First, Translate Later: Interpreting Translation Dynamics in Multilingual Pretraining

• URL: http://arxiv.org/abs/2604.17633v1
• Date: Sun, 19 Apr 2026 22:03:29 GMT
• Title: Copy First, Translate Later: Interpreting Translation Dynamics in Multilingual Pretraining
• Abstract summary: We study the early trajectory of linguistic and translation capabilities in a multilingual 1.7B model.<n>We find that the model quickly acquires basic linguistic capabilities in parallel with token-level copying.<n>We find that translation develops in two distinct phases: an initial phase dominated by copying and surface-level similarities, and a second phase in which more generalizing translation mechanisms are developed.
• Score: 45.58977551849636
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Large language models exhibit impressive cross-lingual capabilities. However, prior work analyzes this phenomenon through isolated factors and at sparse points during training, limiting our understanding of how cross-lingual generalization emerges--particularly in the early phases of learning. To study the early trajectory of linguistic and translation capabilities, we pretrain a multilingual 1.7B model on nine diverse languages, capturing checkpoints at a much finer granularity. We further introduce a novel word-level translation dataset and trace how translation develops over training through behavioral analyses, model-component analysis, and parameter-based ablations. We find that the model quickly acquires basic linguistic capabilities in parallel with token-level copying, while translation develops in two distinct phases: an initial phase dominated by copying and surface-level similarities, and a second phase in which more generalizing translation mechanisms are developed while copying is refined. Together, these findings provide a fine-grained view of how cross-lingual generalization develops during multilingual pretraining.

Related papers

• When Meanings Meet: Investigating the Emergence and Quality of Shared Concept Spaces during Multilingual Language Model Training [57.230355403478995]
  We investigate the development of language-agnostic concept spaces during pretraining of EuroLLM.<n>We find that shared concept spaces emerge early and continue to refine, but that alignment with them is language-dependent.<n>In contrast to prior work, our fine-grained manual analysis reveals that some apparent gains in translation quality reflect shifts in behavior.
  arXiv  Detail & Related papers  (2026-01-30T11:23:01Z)
• Cross-Lingual Generalization and Compression: From Language-Specific to Shared Neurons [20.13484267765109]
  We study how multilingual language models evolve during pre-training.<n>We observe a transition from uniform language identification capabilities across layers to more specialized layer functions.<n>We identify specific neurons that emerge as increasingly reliable predictors for the same concepts across languages.
  arXiv  Detail & Related papers  (2025-06-02T13:06:30Z)
• Analyzing the Mono- and Cross-Lingual Pretraining Dynamics of Multilingual Language Models [73.11488464916668]
  This study investigates the dynamics of the multilingual pretraining process. We probe checkpoints taken from throughout XLM-R pretraining, using a suite of linguistic tasks. Our analysis shows that the model achieves high in-language performance early on, with lower-level linguistic skills acquired before more complex ones.
  arXiv  Detail & Related papers  (2022-05-24T03:35:00Z)
• Cross-lingual Lifelong Learning [53.06904052325966]
  We present a principled Cross-lingual Continual Learning (CCL) evaluation paradigm. We provide insights into what makes multilingual sequential learning particularly challenging. The implications of this analysis include a recipe for how to measure and balance different cross-lingual continual learning desiderata.
  arXiv  Detail & Related papers  (2022-05-23T09:25:43Z)
• Modelling Latent Translations for Cross-Lingual Transfer [47.61502999819699]
  We propose a new technique that integrates both steps of the traditional pipeline (translation and classification) into a single model. We evaluate our novel latent translation-based model on a series of multilingual NLU tasks. We report gains for both zero-shot and few-shot learning setups, up to 2.7 accuracy points on average.
  arXiv  Detail & Related papers  (2021-07-23T17:11:27Z)
• First Align, then Predict: Understanding the Cross-Lingual Ability of Multilingual BERT [2.2931318723689276]
  Cross-lingual transfer emerges from fine-tuning on a task of interest in one language and evaluating on a distinct language, not seen during the fine-tuning. We show that multilingual BERT can be viewed as the stacking of two sub-networks: a multilingual encoder followed by a task-specific language-agnostic predictor. While the encoder is crucial for cross-lingual transfer and remains mostly unchanged during fine-tuning, the task predictor has little importance on the transfer and can be red during fine-tuning.
  arXiv  Detail & Related papers  (2021-01-26T22:12:38Z)
• VECO: Variable and Flexible Cross-lingual Pre-training for Language Understanding and Generation [77.82373082024934]
  We plug a cross-attention module into the Transformer encoder to explicitly build the interdependence between languages. It can effectively avoid the degeneration of predicting masked words only conditioned on the context in its own language. The proposed cross-lingual model delivers new state-of-the-art results on various cross-lingual understanding tasks of the XTREME benchmark.
  arXiv  Detail & Related papers  (2020-10-30T03:41:38Z)
• Translation Artifacts in Cross-lingual Transfer Learning [51.66536640084888]
  We show that machine translation can introduce subtle artifacts that have a notable impact in existing cross-lingual models. In natural language inference, translating the premise and the hypothesis independently can reduce the lexical overlap between them. We also improve the state-of-the-art in XNLI for the translate-test and zero-shot approaches by 4.3 and 2.8 points, respectively.
  arXiv  Detail & Related papers  (2020-04-09T17:54:30Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.